By Jeffrey Zimmet, MD, PhD

Associate Professor of Medicine, University of California, San Francisco; Director, Cardiac Catheterization Laboratory, San Francisco VA Medical Center

Dr. Zimmet reports no financial relationships relevant to this field of study.

SYNOPSIS: This large study suggests that several invasive, nondental medical procedures may be triggers for subsequent infective endocarditis, reopening the debate regarding prevention and management.

SOURCE: Janszky I, Gémes K, Ahnve S, et al. Invasive procedures associated with the development of infective endocarditis. J Am Coll Cardiol 2018;71:2744-2752.

Infective endocarditis (IE) is a condition that occurs in relatively low absolute numbers, but confers a very high risk for morbidity and mortality. U.S. guidelines for the prevention of IE have focused on bacteremia following dental procedures, with less coverage devoted to gastrointestinal (GI) and genitourinary (GU) tract procedures. It seems likely that the 2007 American Heart Association IE guideline update, along with its European counterpart, was well-received by dentists, because it removed the recommendation for prophylactic antibiotics prior to dental work for all but the highest-risk patients (those with a prosthetic heart valve, with a prior history of IE, for certain patients with complex congenital heart disease, and for cardiac transplant recipients with valvulopathy). Likewise, this update removed the recommendation for the use of antibiotic prophylaxis for procedures involving the GI and GU tracts. Other invasive procedures, including so-called “clean” invasive procedures such as coronary angiography, received no mention.

Much of the change in guidelines was based on the lack of convincing evidence for the efficacy of prophylactic antibiotics for prevention of IE. Clinicians often assume that the risk for endocarditis with medical procedures is negligible as well, but data addressing this point are missing from the debate.

To address this shortfall, Janszky et al analyzed all cases of endocarditis in Sweden over a 14-year period following the 1997 adoption of a standardized classification system for coding of medical procedures. To avoid confounding, the authors used a case-crossover design in which each patient served as his or her own control. For each case, the occurrence of medical procedures in the 12-week period preceding the endocarditis diagnosis was compared with the 12-week period one year earlier. Over the course of the study period, 7,013 cases of IE in adult patients were identified. Researchers found multiple invasive procedures were associated with an increased risk of endocarditis. Among outpatient procedures, this included not only GI and GU procedures such as colonoscopy (relative risk [RR], 2.89; 95% confidence interval [CI], 1.35-6.17) and cystoscopy (RR, 1.59; 95% CI, 0.98-2.58), but also coronary angiography (RR, 4.75; 95% CI, 1.61-13.96), bone marrow puncture (RR, 4.33), and bronchoscopy (RR, 5.0), as well as transfusion and hemodialysis. The same procedures performed on an inpatient basis appeared to have similar or stronger associations with subsequent endocarditis, with the RR for bronchoscopy, for example, rising to 16. Coronary artery bypass grafting had an especially strong association (RR, 13.8), as well as a conglomeration of other major and minor cardiovascular therapeutic procedures, including aortic surgery and pacemaker insertion (RR, 9.75). Phacoemulsification, a common procedure that would not be expected to lead to transient bacteremia, was not associated with elevated risk. The study included no information about antibiotic prophylaxis, and the authors did not have access to microbiological data on the pathogens involved in endocarditis.

The authors concluded that multiple invasive medical procedures appear to contribute to the subsequent development of IE. They argued for a potential reconsideration of prophylactic antibiotics for certain high-risk patients and procedures.

However, more strongly, the authors contended that this knowledge supports a renewed focus on aseptic technique in procedures, and that increased awareness of the risk following certain procedures could lead to earlier diagnosis and improved outcomes.


This is the largest study to date linking invasive procedures to an increased risk of endocarditis. The completeness and high reported accuracy of the Swedish National Patient Register add to the strength of the study, and the case-crossover design represents an improvement over traditional case-control studies.

A range of invasive procedures, including but by no means limited to dental procedures, could lead to a transient bacteremia, which is a necessary condition for the formation of an infective vegetation. Transient bacteremia always will be a frequent outcome of certain procedures; cystoscopy and colonoscopy come to mind. However, for others, varying levels of sterile technique can lead to different results. For example, in cardiac catheterization, there is significant variability in sterile technique from institution to institution, with variable use of hats and masks. It is rare that the cath lab is considered a completely sterile environment exactly like the OR. Because of the variable time delay between transient bacteremia and development of clinically evident IE, the inciting procedure may well not be identified as causative on a case-by-case basis. There is room for improved aseptic technique in this procedure, as in others identified in the study, including bone marrow biopsy and basic vascular access for hemodialysis.

The stratified analysis suggested that the risk of invasive inpatient procedures was higher in the latter half of the study period than in the earlier period. Whether this is at all attributable to the newer guidelines restricting prophylactic antibiotics surely will add to the debate. Based on their analysis, the authors estimated that 476 high-risk patients would need to receive prophylactic antibiotics to prevent one case of IE, assuming that prophylaxis was 100% effective. This number was considerably lower for certain high-risk procedures (83 for bronchoscopy, for example). This will add information for future guidelines and, if confirmed, could result in altered prophylaxis recommendations for certain patients and procedures. However, it is more likely that the overall approach will be less about antibiotic prophylaxis and more about improving sterility, where possible, while developing system-based approaches to the management of procedure-related bacteremia.